In biosensors used in blood glucose self-monitoring devices, the value of the current response varies depending on the value of hematocrit in whole blood. It has been an issue that a blood glucose level deviating from its true value may be obtained as a measurement result (hematocrit influence). In view of the hematocrit influence, JP 2013-061336 A discloses a method in which the hematocrit value is measured by applying an alternating voltage in a blood glucose self-monitoring device. The measured hematocrit value is then used to correct the value of the glucose level, so that the deviation from the true value is reduced. Further, JP 4341032 B discloses a method in which a separation membrane is used to separate blood plasma from whole blood, and the blood glucose level in the plasma is then measured in order to prevent the measured glucose level from being affected by the hematocrit value.
However, the method disclosed in JP 2013-061336 A has problems in that the application of alternating voltage increases the electric power consumption by the device, and the accuracy of the blood glucose measurement may be reduced because of overcorrection due to high noise in the measurement of the hematocrit value. Further, in the method disclosed in JP 4341032 B, there are problems in that the structure of a sensor for use in the method may be complex since the sensor requires a separation membrane, and the ability of the separation membrane to separate blood cells may not be stable because of the influence of the viscosity of the sample due to components other than blood cells, such as lipids.